🍄 MycoTrack — Smart Mushroom Farming & AI Health Monitoring System

MycoTrack is an end-to-end IoT and Deep Learning solution designed for smart mushroom farming. It continuously monitors environmental metrics (temperature, humidity, CO₂ levels), captures high-resolution mushroom bed imagery via IoT edge hardware (ESP32-CAM and Raspberry Pi), and runs real-time computer vision diagnostic models (ResNet-50 / YOLOv8) to detect mushroom diseases and growth health.

---

📐 System Architecture

graph TD
    subgraph Hardware Tier
        ESP[ESP32-CAM Module]
        PI[Raspberry Pi Camera Node]
    end

    subgraph Cloud Infrastructure - AWS / Firebase
        IoT[AWS IoT Core / MQTT]
        S3[AWS S3 Bucket: mycotrack-images]
        DDB[(AWS DynamoDB)]
        COG[AWS Cognito Auth]
    end

    subgraph AI Inference Backend
        API[FastAPI Backend - realtime_mushroom.py]
        MODEL[PyTorch Model - ResNet50 / YOLOv8]
    end

    subgraph Client Tier
        APP[MycoTrack Expo React Native App]
    end

    ESP -->|Telemetry & Captures| IoT
    PI -->|Upload Image| S3
    PI -->|Publish Scan Event| IoT
    APP -->|Authenticate| COG
    APP -->|Send Image / Stream| API
    API -->|Inference| MODEL
    API -->|Save Scan Analysis| DDB
    API -->|Upload Captured Frames| S3
    APP -->|Query Telemetry & Scans| DDB
    APP -->|MQTT Control Commands| IoT

Loading

---

📱 Navigation & App Navigation Flow

The mobile app is structured around Expo Router (file-based navigation) located inside app/screens/.

graph TD
    A[Welcome Screen] --> B[Login / Signup]
    B -->|AWS Cognito Auth| C[Dashboard / Home]
    C --> D[Monitoring Screen - Live Telemetry & Actuators]
    C --> E[Camera AI Screen - Real-time Frame Analysis]
    C --> F[Disease Result Screen - Diagnostic Breakdown]
    C --> G[Alerts Screen - Threshold & Outbreak Warnings]
    C --> H[Settings Screen - Hardware & User Preferences]

Loading

🗺 Screen Breakdown

Screen	Route / File	Description
Welcome	app/screens/Welcome.js	Onboarding screen introducing MycoTrack features.
Login / Signup	app/screens/Login.js, Signup.js	AWS Cognito user authentication (Email/Password & Hosted UI).
Dashboard	app/screens/Dashboard.js	Main overview showing recent scans, quick telemetry, and alerts.
Monitoring	app/screens/Monitoring.js	Real-time graphs and controls for temperature, humidity, CO₂ fan & misting actuators.
Camera AI	app/screens/CameraAI.js	Live camera view and file picker for instant AI disease detection.
Disease Result	app/screens/DiseaseResult.js	Detailed AI diagnostic results, confidence scores, and treatment recommendations.
Alerts	app/screens/Alerts.js	Log of environmental threshold alerts and disease warnings.
Settings	app/screens/Settings.js	Configuration for connected IoT devices, notification preferences, and account management.

---

🛠 Frameworks, Libraries & Technologies

1. Mobile App

• Core Framework: Expo ~54.0.33 with React Native 0.81.5 and React 19.1.0
• Routing: expo-router ~6.0.23 (File-based app routing)
• Styling: styled-components ^6.3.12, expo-linear-gradient, @expo/vector-icons
• Authentication: AWS Amplify (@aws-amplify/auth, @aws-amplify/react-native)
• Cloud & Database: AWS SDK for JS (@aws-sdk/client-dynamodb, @aws-sdk/lib-dynamodb, @aws-sdk/client-iot-data-plane), firebase (legacy integration)
• Realtime IoT: mqtt ^5.15.1
• Device Capabilities: expo-camera, expo-image-picker, expo-notifications, expo-haptics

2. AI Inference Backend (backend/)

• API Framework: FastAPI & Uvicorn ASGI Server
• Deep Learning Engine: PyTorch (torch, torchvision) with trained weights (mushroom_cls.pth)
• Architectures Supported: ResNet-50 Classifier & YOLOv8 Detection Pipeline
• Image Processing: OpenCV (opencv-python-headless), Pillow (PIL)
• Cloud Connectivity: boto3 (AWS S3, DynamoDB, IoT Data Plane)

3. Hardware & Firmware

• ESP32-CAM: Arduino sketch (esp32_cam_aws/esp32_cam_aws.ino) streaming MQTT telemetry and JPEG snapshots over SSL.
• Raspberry Pi: Python automation scripts (src/AWS/pi_capture_aws.py) for automated scheduled image captures and AWS S3/DynamoDB sync.

---

🗄 Database Schema (AWS DynamoDB)

• MycoTrack_Scans:
  • userId (String, Partition Key)
  • createdAt (Number, Sort Key)
  • imageUrl (S3 URL), diseaseResult (Healthy/Unhealthy), confidence (Float), boundingBoxes
• MycoTrack_Alerts:
  • houseId (String, Partition Key)
  • alertId (String, Sort Key)
  • type (Temperature/Humidity/Disease), severity, timestamp
• MycoTrack_Sensors:
  • houseId (String, Partition Key)
  • timestamp (Number, Sort Key)
  • temperature, humidity, co2

---

🚀 How to Run the Project

Prerequisites

• Node.js (v18+ recommended)
• Python 3.9+ (for backend)
• Expo Go App on your mobile device (or Android Studio / Xcode for emulators)

---

Step 1: Run the Mobile App (Expo)

From the root directory:

1. Navigate to the my-app directory:

   cd my-app

2. Install node dependencies (if not already installed):

   npm install

3. Start the Expo server:

   npx expo start

4. Launch the App:

   • On Mobile: Scan the displayed QR code using the Expo Go app (Android) or Camera app (iOS).
   • In Web Browser: Press w in the terminal to launch the web client.
   • On Android Emulator: Press a in the terminal.
   • On iOS Simulator: Press i in the terminal.

---

Step 2: Run the AI Backend Server (FastAPI)

1. Navigate to the backend directory:

   cd my-app/backend

2. Activate the Python Virtual Environment:

   • Windows:

     .\venv\Scripts\activate

   • macOS / Linux:

     source venv/bin/activate

3. Install python packages (if needed):

   pip install fastapi uvicorn torch torchvision pillow boto3 opencv-python-headless

4. Start the FastAPI server:

   python -m uvicorn realtime_mushroom:app --host 0.0.0.0 --port 8000 --reload

5. Access API Documentation: Open http://localhost:8000/docs in your browser to view and test interactive Swagger API endpoints.